Absolute temperature reading

[evertheylen]

I know this is on the TODO list, but I would like to open an issue for it to get more details. I would like to use this in a project where an absolute reading would be very useful, and the Thermal Seek looks to be the best solution, using your library. So I'd like to ask, is there perhaps any progress on this? Can I help? I'll buy the Thermal Seek myself soon.

[zougloub]

Hi, opening the issue is perfectly fine :)

So far I have done absolutely no progress as I'm fairly busy with my day job. I'm pretty sure that (unless obfuscation is done) given the technology used by the camera, it shouldn't be impossible to get absolute readings, by trying to get a model of the camera (from how it works) and finding out where the required information comes from, in the data stream.

If you want to help, you can look in the related links to check if people are making progress (last time I checked, people were doing very approximative relative temperature readings) and you could try to have fun and think about how the camera works.

Building a model can help finding out to properly correct fixed pattern noise and other imaging defects (other TODOs).

[zougloub]

Hi,

I am reusing my device and took another look at absolute temperature measurement:

• I took a look at pixels with low standard deviation, and pixel (0,1) (stream offset 2) was outstanding.
• I noticed that when the camera temperature raises, pixel (0,1) value lowers.

It seems there is a "thermal diode" in the device, if someone can find pointers about it, maybe we can go further and convert the value into a temperature.